Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Doug Evans <dje@google.com>
To: Pedro Alves <pedro@codesourcery.com>, Ian Lance Taylor <iant@google.com>
Cc: Stan Shebs <stan@codesourcery.com>,
	gdb-patches@sourceware.org,        Eli Zaretskii <eliz@gnu.org>,
	tromey@redhat.com
Subject: Re: [NEWS/RFA] Re: [gdbserver] x86 agent expression bytecode compiler 	(speed up conditional tracepoints)
Date: Wed, 16 Jun 2010 19:21:00 -0000	[thread overview]
Message-ID: <AANLkTimcnED-loxkksN4lm7St5DzhQUjdbl4hYO7he9_@mail.gmail.com> (raw)
In-Reply-To: <201006162016.18181.pedro@codesourcery.com>

On Wed, Jun 16, 2010 at 12:16 PM, Pedro Alves <pedro@codesourcery.com> wrote:
> On Wednesday 16 June 2010 20:12:44, Stan Shebs wrote:
>> Doug Evans wrote:
>> > On Mon, Jun 14, 2010 at 3:19 PM, Pedro Alves <pedro@codesourcery.com> wrote:
>> >
>> >> Thanks.  I've checked the whole thing in.
>> >>
>> >
>> > I'm getting build failures that go away when compiling linux-x86-low.c with -g.
>> >
>> > gcc is optimizing out the skipped over stuff I believe.
>> >
>>
>> Hmm, this particular trickery was my idea, but it always seemed
>> vulnerable to ever-smarter optimization.  Does anybody have any better
>> strategy?  Declaring the asm volatile didn't work, because the compiler
>> was whacking everything around it.
>>
>> The other approach is to build up instructions from bitfields, which is
>> reliable but needs a lot of setup and helper routines.
>
> Quick thought: can we stick a couple of __attribute__((used))'s in the
> macro, so the compiler doesn optimized things away, thinking they're
> unused (given the uses are behind asm)?
>
>>
>> Stan
>> >
>> > /* Our general strategy for emitting code is to avoid specifying raw
>> >    bytes whenever possible, and instead copy a block of inline asm
>> >    that is embedded in the function.  This is a little messy, because
>> >    we need to keep the compiler from discarding what looks like dead
>> >    code, plus suppress various warnings.  */
>> >
>> > #define EMIT_ASM(NAME,INSNS)                                            \
>> >   { extern unsigned char start_ ## NAME, end_ ## NAME;                  \
>> >     add_insns (&start_ ## NAME, &end_ ## NAME - &start_ ## NAME);       \
>> >     if (always_true ())                                         \
>> >       goto skipover ## NAME;                                            \
>> >     __asm__ ("start_" #NAME ":\n\t" INSNS "\n\tend_" #NAME ":\n\t");    \
>> >     skipover ## NAME:                                                   \
>> >     ; }

This worked with the gcc4.4.0 variant I was using.

Ian: Is there a Right way to do this?

2010-06-16  Doug Evans  <dje@google.com>

        * linux-x86-low.c (always_true): Change to global.
        (EMIT_ASM, EMIT_ASM32): Update.

Index: linux-x86-low.c
===================================================================
RCS file: /cvs/src/src/gdb/gdbserver/linux-x86-low.c,v
retrieving revision 1.19
diff -u -p -r1.19 linux-x86-low.c
--- linux-x86-low.c     15 Jun 2010 10:44:48 -0000      1.19
+++ linux-x86-low.c     16 Jun 2010 19:15:21 -0000
@@ -1484,13 +1484,8 @@ add_insns (unsigned char *start, int len
   current_insn_ptr = buildaddr;
 }

-/* A function used to trick optimizers.  */
-
-int
-always_true (void)
-{
-  return 1;
-}
+/* Used to trick optimizers.  */
+static volatile int always_true = 1;

 /* Our general strategy for emitting code is to avoid specifying raw
    bytes whenever possible, and instead copy a block of inline asm
@@ -1501,7 +1496,7 @@ always_true (void)
 #define EMIT_ASM(NAME,INSNS)                                           \
   { extern unsigned char start_ ## NAME, end_ ## NAME;                 \
     add_insns (&start_ ## NAME, &end_ ## NAME - &start_ ## NAME);      \
-    if (always_true ())                                                \
+    if (always_true)                                           \
       goto skipover ## NAME;                                           \
     __asm__ ("start_" #NAME ":\n\t" INSNS "\n\tend_" #NAME ":\n\t");   \
     skipover ## NAME:                                                  \
@@ -1513,7 +1508,7 @@ always_true (void)
 #define EMIT_ASM32(NAME,INSNS)                                         \
   { extern unsigned char start_ ## NAME, end_ ## NAME;                 \
     add_insns (&start_ ## NAME, &end_ ## NAME - &start_ ## NAME);      \
-    if (always_true ())                                                \
+    if (always_true)                                           \
       goto skipover ## NAME;                                           \
     __asm__ (".code32\n\tstart_" #NAME ":\n\t" INSNS "\n\tend_" #NAME ":\n" \
             "\t.code64\n\t");                                          \


  reply	other threads:[~2010-06-16 19:21 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-06-07 16:00 Pedro Alves
2010-06-07 16:04 ` Michael Snyder
2010-06-07 16:35   ` Joel Brobecker
2010-06-10 17:28 ` Tom Tromey
2010-06-10 17:36   ` Pedro Alves
2010-06-10 18:43     ` Tom Tromey
2010-06-14 11:15       ` [NEWS/RFA] " Pedro Alves
2010-06-14 17:29         ` Eli Zaretskii
2010-06-14 22:19           ` Pedro Alves
2010-06-16 18:57             ` Doug Evans
2010-06-16 19:13               ` Stan Shebs
2010-06-16 19:16                 ` Pedro Alves
2010-06-16 19:21                   ` Doug Evans [this message]
2010-06-16 23:58                     ` Ian Lance Taylor
2010-06-20 22:27                       ` Pedro Alves
2010-06-16 19:31                   ` Stan Shebs
2010-06-17 18:03                     ` Michael Snyder
2010-06-19 16:13                       ` Hui Zhu
2010-06-19 16:16                         ` Doug Evans
2010-06-19 17:13                           ` Hui Zhu
2010-06-19 17:26                             ` Doug Evans
2010-06-20  9:30                               ` [RFA-new version][gdbserver] " Pierre Muller
     [not found]                               ` <-3945058798826177264@unknownmsgid>
2010-06-20 15:30                                 ` Doug Evans
2010-06-20 17:02                                   ` Pierre Muller
     [not found]                                   ` <-1673004315710326113@unknownmsgid>
2010-06-20 17:11                                     ` Doug Evans
2010-06-20 18:41                                   ` Tom Tromey
2010-06-20 20:36                                     ` Pierre Muller
2010-06-20 21:07                                       ` Pierre Muller
2010-06-21  1:47                                         ` Tom Tromey
2010-06-21  6:31                                           ` Pierre Muller
2010-06-20 18:31                         ` [NEWS/RFA] Re: [gdbserver] " Tom Tromey

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=AANLkTimcnED-loxkksN4lm7St5DzhQUjdbl4hYO7he9_@mail.gmail.com \
    --to=dje@google.com \
    --cc=eliz@gnu.org \
    --cc=gdb-patches@sourceware.org \
    --cc=iant@google.com \
    --cc=pedro@codesourcery.com \
    --cc=stan@codesourcery.com \
    --cc=tromey@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox